jpmschweitzer/settled-reach
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

fix(ci): address PR #35 review comments

Browse Source 
- Document non-blocking receive contract in perf_bench.rs docstring,
  confirming no TCP deadlock race (Hoshe #1, critical)
- Make shadowcast parser order-independent — flush on new config header
  instead of requiring Recursive after Symmetric (Hoshe #2)
- Fix p95 calculation: use floor(0.95*(N-1)) nearest-rank instead of
  ceil(0.95*N)-1 which was off-by-one at N=50 (Hoshe #4)
- Error on --compare when no baseline file exists (Hoshe #5)
- Add D-031 10tps assumption comment to TICK_BUDGET_US (Tyre #1)
- Strengthen snapshot assertion: require warmup + half measurement
  window instead of warmup + 1 (Tyre #2)
- Regenerate baseline with corrected p95 (356µs, was 526µs)

Hoshe #3 (.PHONY) was already addressed — perf-baseline is in the
.PHONY declaration on Makefile line 10.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
This commit is contained in:
jpmschweitzer
2026-02-18 13:07:46 +01:00
co-authored by Claude Opus 4.6
parent c1d7c072ae
commit d0596d2763
3 changed files with 72 additions and 52 deletions
Download Patch File Download Diff File Expand all files Collapse all files
server/tests/perf_bench.rs
+16 -5
View File
@@ -55,6 +55,12 @@ fn read_rss_kb() -> Option<u64> {
/// Boots the server with production content, runs WARMUP_TICKS to stabilize, /// Boots the server with production content, runs WARMUP_TICKS to stabilize,
/// then measures MEASURE_TICKS of app.update() wall-clock time. Reports entity /// then measures MEASURE_TICKS of app.update() wall-clock time. Reports entity
/// counts from observer snapshots and process RSS. /// counts from observer snapshots and process RSS.
///
/// Protocol contract: BridgePlugin uses non-blocking receive (WouldBlock →
/// empty input vec), so the server always advances even if the client hasn't
/// sent input yet. The server sends a snapshot each tick; the client blocks
/// on read until one arrives, then responds with (empty) input. No deadlock
/// possible — see TcpBridge::receive_inputs and send_snapshot in bridge/tcp.rs.
#[test] #[test]
#[ignore] #[ignore]
fn perf_tick_timing() { fn perf_tick_timing() {
@@ -145,12 +151,16 @@ fn perf_tick_timing() {
} }
} }
// Must have received enough data past warmup for meaningful results // Must have received enough measured snapshots for meaningful results.
// Require all warmup ticks plus at least half the measurement window.
let min_snapshots = WARMUP_TICKS + MEASURE_TICKS / 2;
assert!( assert!(
entity_counts.len() > WARMUP_TICKS, entity_counts.len() >= min_snapshots,
"Only received {} snapshots, need at least {} (warmup) + 1", "Only received {} snapshots, need at least {} ({} warmup + {} measured)",
entity_counts.len(), entity_counts.len(),
WARMUP_TICKS min_snapshots,
WARMUP_TICKS,
MEASURE_TICKS / 2
); );
drop(reader); drop(reader);
@@ -174,7 +184,8 @@ fn perf_tick_timing() {
let mut sorted = measured_us.clone(); let mut sorted = measured_us.clone();
sorted.sort(); sorted.sort();
let p95_idx = ((sorted.len() as f64 * 0.95).ceil() as usize).saturating_sub(1); // Nearest-rank p95: index = floor(0.95 * (N-1)) for 0-based indexing.
let p95_idx = ((sorted.len() - 1) as f64 * 0.95).floor() as usize;
let p95 = sorted[p95_idx.min(sorted.len() - 1)]; let p95 = sorted[p95_idx.min(sorted.len() - 1)];
let entity_counts_measured: Vec<usize> = let entity_counts_measured: Vec<usize> =
tests/perf/baseline.json
+43 -43
View File
@@ -1,15 +1,15 @@
{ {
"timestamp": "2026-02-18T11:39:46.913254+00:00", "timestamp": "2026-02-18T12:07:18.546680+00:00",
"git": { "git": {
"commit": "b4a3784", "commit": "c1d7c07",
"branch": "ci" "branch": "ci"
}, },
"tick_timing": { "tick_timing": {
"max_us": 537, "max_us": 363,
"mean_us": 366, "mean_us": 332,
"measured_ticks": 50, "measured_ticks": 50,
"min_us": 314, "min_us": 310,
"p95_us": 526, "p95_us": 356,
"warmup_ticks": 5 "warmup_ticks": 5
}, },
"entities": { "entities": {
@@ -17,7 +17,7 @@
"max_per_snapshot": 21 "max_per_snapshot": 21
}, },
"memory": { "memory": {
"rss_kb": 6972 "rss_kb": 6900
}, },
"shadowcast": { "shadowcast": {
"configs": [ "configs": [
@@ -26,90 +26,90 @@
"density": "open field", "density": "open field",
"range": 20, "range": 20,
"iterations": 1000, "iterations": 1000,
"symmetric_total_ms": 62.2, "symmetric_total_ms": 57.86,
"symmetric_per_call_us": 62.2, "symmetric_per_call_us": 57.86,
"recursive_total_ms": 107.25, "recursive_total_ms": 97.97,
"recursive_per_call_us": 107.25 "recursive_per_call_us": 97.97
}, },
{ {
"map_size": 32, "map_size": 32,
"density": "moderate corridors", "density": "moderate corridors",
"range": 20, "range": 20,
"iterations": 1000, "iterations": 1000,
"symmetric_total_ms": 60.15, "symmetric_total_ms": 52.29,
"symmetric_per_call_us": 60.15, "symmetric_per_call_us": 52.29,
"recursive_total_ms": 211.61, "recursive_total_ms": 189.99,
"recursive_per_call_us": 211.61 "recursive_per_call_us": 189.99
}, },
{ {
"map_size": 32, "map_size": 32,
"density": "dense rooms", "density": "dense rooms",
"range": 20, "range": 20,
"iterations": 1000, "iterations": 1000,
"symmetric_total_ms": 24.28, "symmetric_total_ms": 21.31,
"symmetric_per_call_us": 24.28, "symmetric_per_call_us": 21.31,
"recursive_total_ms": 107.95, "recursive_total_ms": 98.47,
"recursive_per_call_us": 107.95 "recursive_per_call_us": 98.47
}, },
{ {
"map_size": 64, "map_size": 64,
"density": "open field", "density": "open field",
"range": 20, "range": 20,
"iterations": 1000, "iterations": 1000,
"symmetric_total_ms": 75.38, "symmetric_total_ms": 60.5,
"symmetric_per_call_us": 75.38, "symmetric_per_call_us": 60.5,
"recursive_total_ms": 108.22, "recursive_total_ms": 98.33,
"recursive_per_call_us": 108.22 "recursive_per_call_us": 98.33
}, },
{ {
"map_size": 64, "map_size": 64,
"density": "moderate corridors", "density": "moderate corridors",
"range": 20, "range": 20,
"iterations": 1000, "iterations": 1000,
"symmetric_total_ms": 48.72, "symmetric_total_ms": 47.41,
"symmetric_per_call_us": 48.72, "symmetric_per_call_us": 47.41,
"recursive_total_ms": 202.49, "recursive_total_ms": 184.87,
"recursive_per_call_us": 202.49 "recursive_per_call_us": 184.87
}, },
{ {
"map_size": 64, "map_size": 64,
"density": "dense rooms", "density": "dense rooms",
"range": 20, "range": 20,
"iterations": 1000, "iterations": 1000,
"symmetric_total_ms": 14.5, "symmetric_total_ms": 14.35,
"symmetric_per_call_us": 14.5, "symmetric_per_call_us": 14.35,
"recursive_total_ms": 90.33, "recursive_total_ms": 79.6,
"recursive_per_call_us": 90.33 "recursive_per_call_us": 79.6
}, },
{ {
"map_size": 150, "map_size": 150,
"density": "open field", "density": "open field",
"range": 20, "range": 20,
"iterations": 1000, "iterations": 1000,
"symmetric_total_ms": 59.7, "symmetric_total_ms": 56.65,
"symmetric_per_call_us": 59.7, "symmetric_per_call_us": 56.65,
"recursive_total_ms": 103.09, "recursive_total_ms": 97.8,
"recursive_per_call_us": 103.09 "recursive_per_call_us": 97.8
}, },
{ {
"map_size": 150, "map_size": 150,
"density": "moderate corridors", "density": "moderate corridors",
"range": 20, "range": 20,
"iterations": 1000, "iterations": 1000,
"symmetric_total_ms": 43.64, "symmetric_total_ms": 40.49,
"symmetric_per_call_us": 43.64, "symmetric_per_call_us": 40.49,
"recursive_total_ms": 178.09, "recursive_total_ms": 158.81,
"recursive_per_call_us": 178.09 "recursive_per_call_us": 158.81
}, },
{ {
"map_size": 150, "map_size": 150,
"density": "dense rooms", "density": "dense rooms",
"range": 20, "range": 20,
"iterations": 1000, "iterations": 1000,
"symmetric_total_ms": 12.63, "symmetric_total_ms": 12.88,
"symmetric_per_call_us": 12.63, "symmetric_per_call_us": 12.88,
"recursive_total_ms": 79.87, "recursive_total_ms": 77.98,
"recursive_per_call_us": 79.87 "recursive_per_call_us": 77.98
} }
] ]
} }
tooling/perf-baseline
+13 -4
View File
@@ -22,7 +22,8 @@ ROOT = Path(__file__).resolve().parent.parent
PERF_DIR = ROOT / "tests" / "perf" PERF_DIR = ROOT / "tests" / "perf"
BASELINE_FILE = PERF_DIR / "baseline.json" BASELINE_FILE = PERF_DIR / "baseline.json"
# Tick budget from D-026 # Tick budget from D-026: 100ms per tick at 10 tps floor (D-031).
# If tick rate changes, update this constant.
TICK_BUDGET_US = 100_000 # 100ms TICK_BUDGET_US = 100_000 # 100ms
@@ -95,6 +96,9 @@ def run_shadowcast_benchmark():
line, line,
) )
if m: if m:
# New config block — flush previous if complete
if current.get("map_size"):
configs.append(current)
current = { current = {
"map_size": int(m.group(1)), "map_size": int(m.group(1)),
"density": m.group(3), "density": m.group(3),
@@ -113,11 +117,12 @@ def run_shadowcast_benchmark():
if m: if m:
current["recursive_total_ms"] = float(m.group(1)) current["recursive_total_ms"] = float(m.group(1))
current["recursive_per_call_us"] = float(m.group(2)) current["recursive_per_call_us"] = float(m.group(2))
if current.get("map_size"):
configs.append(current)
current = {}
continue continue
# Flush last config
if current.get("map_size"):
configs.append(current)
return {"configs": configs} if configs else None return {"configs": configs} if configs else None
@@ -252,6 +257,10 @@ def main():
if compare_mode and regressions: if compare_mode and regressions:
print(f"\n{len(regressions)} regression(s) detected.") print(f"\n{len(regressions)} regression(s) detected.")
return 1 return 1
elif compare_mode:
print(f"\nERROR: --compare requires a saved baseline at {BASELINE_FILE.relative_to(ROOT)}")
print("Run `make perf-baseline` first to create one.")
return 1
if compare_mode: if compare_mode:
return 0 return 0